Study on the multi-modal data preprocessing for knowledge-converged super brain

Due to the ever-growing amount of data produced through versatile hyper-connected networks, in pursuit of the Internet of Everything (IoE), the challenges to derive useful knowledge from these big IoE data have received increasing attention. This research deals with fusing multi-modal data from the IoE networks, in order to efficiently provide a set of highly relevant input data for further high-level data analytics such as machine learning. We propose a functional structure of such a data fusion and implement a system prototype based on Apache Spark. Further, our concept is validated through the experiments using the KDD Cup 1999 data.